Mitochondria are semiautonomous organelles located in the cytoplasm. Unlike other organelles, mitochondria have their own genomes (mtDNAs), double-stranded circular DNAs, and are coated by a double-layered membrane (Davis and Williams, 2012; van der Bliek et al., 2017). Mitochondria are responsible for the production of adenosine triphosphate (ATP) and the regulation of nutritional metabolism, calcium homeostasis, and cellular viability in several organs (Kim et al., 2008). Due to the huge demand for ATP in the heart, mitochondria are highly abundant in cardiac cells, especially in cardiomyocytes (CMs), occupying nearly 20–40% volume in adult CMs (Schaper et al., 1985). Approximately 6 kg of ATP is produced by cardiac mitochondria every day (Siasos et al., 2018). Thus, mitochondria play essential roles in the cardiovascular system.

The main source of energy obtained by oxidative phosphorylation (OXPHOS) comes from the oxidation of fatty acids in the adult heart (Stanley, 2005). ATP is synthesized in adult hearts following the steps below: fatty acyl-coenzyme A (CoAs) are synthesized with the help of acyl CoA syntheses. To enter the cardiac mitochondria, fatty acyl CoAs are converted to acylcarnitines by carnitine palmitoyl transferase 1 (CPT1) and are transferred to the inner membrane of mitochondria (IMM) where they are liberated to fatty acyl CoAs and initiate β-oxidation. Acetyl CoA, a product of β-oxidation, enters the Kerbs cycle and generates reduced nicotinamide adenine dinucleotide (NADH) and reduced flavin adenine dinucleotide (FADH₂). NADH and FADH₂ generated from the Kerbs cycle, and from β-oxidation, transfer electrons through the electron transport chain (ETC), which comprises four complexes on the IMM. The efflux of protons accompanied by the electronic flow on ETC activates ATP-synthase and produces ATP (shown in Figure 1) (Chen and Butow, 2005).

However, the production of ATP through the respiratory chain is accompanied by an inevitable generation of ROS, especially superoxide anion (·O₂ ⁻). It has been reported that nearly 1, 2% of electrons is related to superoxide production during ATP synthesis (Boveris et al., 1976). Complex I and III are the main sites of ROS production in mitochondria (Cadenas, 2018). Complex I is an L-shaped component made up of a hydrophilic arm and a hydrophobic arm, in which the semi-/reduced flavin mononucleotide (FMN) and N1a/N1b iron-sulfur clusters are the main sites for superoxide production (Efremov and Sazanov, 2011). Complex III generates superoxide mostly by ubisemiquinone formed at the Q_o site, where ubisemiquinone transfers electrons to oxygen (Iwata et al., 1998). ROS generated from mitochondrial complex I and III is the main mechanism of ischemia/reperfusion (I/R) injury in the heart. Kang et al. reported that complex I is responsible for the generation of O₂ ⁻ in post-ischemic CMs, while damage to the N1a cluster causes FMN-produced superoxide (Kang et al., 2018). Meanwhile, Chen et al. found an increasing generation of antimycin A-enhanced superoxide in complex III ex vivo (Chen et al., 2008). Normal amounts of O₂ ⁻ can be converted to more stable hydrogen peroxide by superoxide dismutases (SODs) (Suzuki et al., 2013). However, excessive O₂ ⁻ can damage mtDNA, lipids, and proteins, leading to mitochondrial dysfunction, oxidative stress, and cell death (Haworth and Hunter, 1979; Peoples et al., 2019). In cardiovascular diseases (CVDs), it has been reported that excessive ROS and mitochondrial dysfunction are related to numerous cardiac diseases, such as atherosclerosis (AS), I/R injury, cardiac hypertrophy, heart failure (HF), and degenerative aortic valve disease (Peoples et al., 2019).

The classical perception is that ROS is the major driver of the mutations of mtDNA, while others are replicated errors of mtDNA, and present a less efficient mtDNA repair system than nuclear DNA. mtDNA, which is closed to ROS and ETC, causes the degradation of nucleic acid chemical components. The most common induction of mutations is via point mutation of purines and pyrimidines (Lee and Wei, 2007). An accumulation of multiple mutations leads to the exhaustion of mtDNA-encoded proteins and less efficient ATP production. It is also recognized that accumulated mtDNA damage can also increase ROS generation, resulting in a vicious circle (Płoszaj et al., 2010). Additionally, mitochondria trigger and sustain the chronic-sterile inflammation and immunity in the heart, which mostly occur after I/R damage (shown in Figure 2), and are related to the development of HF (Nakayama and Otsu, 2018). This phenomenon is induced by activated pattern recognition receptors (PRRs, e.g., toll-like receptor 9 [TLR9] and nod-like receptor pyrin domain containing 3 [NLRP3]) sensing damage-associated molecular patterns (DAMPs) (Rodrigues et al., 2015). As for the characteristics of mtDNA (close to the respiratory chain, sensitive to ROS and a relatively low level of methylation), the accumulation of mtDNA mutations represents the most well-known DAMP, which activates PRRs, recruits inflammatory cytokines, and activates the response of immune cells in the heart.

Mechanisms of mtDNA release from mitochondria remain controversial. The classical theory believes that the mtDNA release is mediated by the opening of the mitochondrial permeability transition pore (MPTP) (Patrushev et al., 2004). The MPTP is an unspecific pore in the mitochondria initially found by Haworth et al. (Haworth and Hunter, 1979). MPTP is strongly related to acute I/R injury and is open under conditions of oxidative stress, overloaded [Ca²⁺]_mito and altered phosphate level during ischemia (Crompton et al., 1987). Although the opening of the MPTP is not ROS-dependent, excessive ROS synergizes the opening of the MPTP with other factors (mentioned above), thereby leading to further mitochondrial injury (Haworth and Hunter, 1979). In addition, evidence has shown that inhibiting the MPTP is an essential mechanism of cardioprotective ischemic pre-/post-conditioning (iPre and iPost) in the heart (reviewed by Penna et al.) (Hausenloy et al., 2009). Opening of the MPTP causes the swelling of the mitochondrial matrix and damage to the outer mitochondrial membrane (OMM), therefore leading to the release of pro-apoptotic factors. Especially, the opening of the MPTP is the major way of releasing mtDNAs from mitochondria and induces sterile inflammation. Therefore, the MPTP is recognized as a promising target against mitochondrial dysfunction in the heart, and several studies from animal models and clinical trials have investigated this target (details are in Section 3.4).

It is also reported that mtDNA release is a voltage-dependent anion channel (VDAC) oligomers-regulated process. Mitochondrial outer membrane permeabilization (MOMP) is required for the release of the mtDNA. Kim et al. found that VDAC can oligomerize under the stimulation of oxidative stress, and induce the MOMP as well as the liberation of mtDNA (Kim et al., 2019). Riley et al. reported that the release of the mtDNA is also related to the mitochondrial inner membrane permeabilization (MIMP). A Bax/BAK (pro-apoptotic member B-cell lymphoma 2, Bcl-2 family)-induced expansion of OMM pores following behind MOMP causes extrusion of IMM from mitochondrion to cytoplasm and the occurrence of MIMP-mediated mtDNA release (Riley et al., 2018). Besides, mitochondrial-derived vesicles (MDVs) have been described as an additional regulator of mitochondrial homeostasis (the major mechanisms are mentioned in Section 2), which establish crosstalk between mitochondria and lysosomes. Dysfunctional mitochondria could lead to leakages of MDVs including their components, which trigger PRRs and activate inflammation. However, the hypothesis of MDVs-dependent mtDNA release remains under investigation (Picca et al., 2019; Picca et al., 2021).

In the next step, mutant mtDNA can be released into intracellular and extracellular spaces. For intracellular liberation, damaged mtDNAs released from disruptive OMM to the cytoplasm, which activates intracellular PRRs and initiates the non-infectious inflammation via interleukin 10 (IL-10) and IL-1β. mtDNA can also induce the activation of cyclic GMP–AMP synthetase (cGAS)/stimulator of interferon gene (STING)/type I interferon (IFN) pathways without inflammatory cytokines activation, especially in the I/R injured heart (White et al., 2014). Several pieces of evidence support the statement that mtDNA can also release out of cells, activating inflammation and stimulating an immune response. In necrotic cells, mtDNA leaks through the disruptive plasma membrane into extracellular space. There are some other mechanisms for the liberation of mtDNA from non-necrotic cells. Guescini et al. detected the existence of cell-free mtDNA in exosomes released by glioblastoma cells and astrocytes, which implies exosomes proceed with the extracellular release of mtDNA (Guescini et al., 2010). Lood et al. reported the release of neutrophils extracellular traps (NETs) is also accompanied by the liberation of mtDNA (Lood et al., 2016). Extracellular leakages of mtDNAs can be recognized by TLR9 through endocytosis or directly captured by TLR9 on the plasma membrane of some cells (e.g., resting B cells and peripheral blood monocytes) and activate inflammation (Eaton-Bassiri et al., 2004; Dasari et al., 2005).

It is not surprising to see that accumulated mtDNA damages and mtDNA-induced chronic inflammation are correlated to AS, I/R injury, hypertension and HF due to the high consumption of oxygen and ATP in the heart (Poznyak et al., 2020). Among them, AS is strongly related to mtDNA mutations. Primary AS (AS that cannot be explained by classical risk factors, such as diabetes, hyperlipidemia, and smoking) has been described as a result of mitochondrial disorder. A study analyzed twelve lipofibrous plaques of aorta intima compared to intact area, and reported four mtDNA mutations, including m.1555A > G, m.3256C > T, m.12315G > A and m.15059G > A, are relevant to atherosclerotic plaques (Sobenin et al., 2013). An analysis enrolled 65 blood samples and 23 atherosclerotic plaques from coronary artery disease (CAD) patients showed a higher deletion of 4977bp (26.2% vs 4.5%) as well as higher heteroplasmy rates (18–46%) than the control group. Interestingly, these deletions are independent of classical risk factors of AS and not inferred by baseline characteristics of patients (Sobenin et al., 2013). Furthermore, Sazonovo et al. showed the data from 225 myocardial infarction (MI) patients, that a mutation of m.5178C > A is related to the occurrence of MI. On the opposite, the other two variants of m.14846G > A and m.12315G > A are observed in healthy controls (Sazonova et al., 2018). A syndrome of Mitochondrial Encephalomyopathy, Lactic Acidosis, and at least one Stroke-like episode (MELAS) is one of the representative mitochondrial diseases carrying a mutant of m.3243A > G in mitochondrial tRNA leucine 1 (Manwaring et al., 2007). Numerous studies found patients with MELAS are also accompanied by dysfunctional endothelia, atherosclerotic carotid, and cerebral artery without potential risk factors of AF. These studies also provide indirect proof of the development of mutant mtDNA-related AS (Pek et al., 2019; Finsterer, 2020). Besides, mtDNA damages also play an essential role in the development of AS accelerated by traditional risk factors mentioned above. A review from Sobenin et al. well summarized the variants of mtDNA relevant to smoking, MELAS, and diabetes mellitus, which are the promoters of the development of AS (Sobenin et al., 2015).

On the other hand, chronic inflammatory and immune responses induced by mtDNA-related DAMPs are also involved in AS (Sobenin et al., 2015). mtDNA mutations affect many cell types in AS. In endothelial cells (ECs), mtDNA activates TLR9 and induces the production of interleukin-1β (IL-1β) through the NF-κB pathway. Moreover, the damage associated with ECs recruits inflammatory cells and alters vascular wall permeability, which initiates the development of AS (Jin et al., 2019). The release of mtDNA induces the activation of NF-κB in macrophages and is related to the synthesis of pro-inflammatory cytokines, uptake of oxidized low-density lipoprotein (oxLDL), and formation of foam cells (Tedgui and Mallat, 2006). Dysfunction of mitochondrial dynamics induced by damaged mtDNA in smooth muscle cells (SMCs) results in their energetic decline and apoptosis in plaque, which alters the stability of plaque in AS (Yu et al., 2017; Docherty et al., 2018). Besides, the release of mutant mtDNA from damaged mitochondria to the cytoplasm is a sign of the early phase of AS. Yu et al. detected mtDNA in both circulating and vascular cells in human atherosclerotic plaques even before other signs of AS occur (Yu et al., 2013). Together all, mtDNA mutations and mtDNA-induced sterile inflammations are essential targets/aspects for the detection and treatment of AS.

To avoid the accumulation of ROS and mitochondrial dysfunction, mitochondria sustain their homeostasis by mitochondrial biogenesis, mitochondrial dynamics, and mitophagy (summarized in Figure 3), all of which are potential targets for treatment.

Mitochondrial biogenesis increases the number of healthy mitochondria in the heart, improves the replication and repair of mtDNA, and induces the synthesis of mitochondrial enzymes and proteins (Popov, 2020). It has been reported that mitochondrial transcription factor A (TFAM) and peroxisome proliferator-activated receptor-γ coactivator-1α (PGC-1α) are central regulators of mitochondrial biogenesis (Jornayvaz and Shulman, 2010). PGC-1α activates TFAM, which is essential for the transcription, replication, and stability of mtDNA (Yokokawa et al., 2018). PGC-1α can also increase the expression of nuclear respiratory factor (NRF)-1 and NRF-2, thereby promoting the production of vital mitochondrial enzymes (Reznick et al., 2007). There are several pathways contributing to mitochondrial biogenesis. Jager et al. reported that adenosine 5‘-monophosphate-activated protein kinase (AMPK) is critical for mitochondrial biogenesis by regulating the phosphorylation of Thr177 and Ser538 (two sites of PGC1α) (Jäger et al., 2007). Sirtuins (SIRT) are a series of deacetylases and deacetylases that control energy metabolism. There is evidence that SIRT1 promotes the uptake of glucose and mitochondrial biogenesis by activating PGC-1α (Guarente, 2007). Wang et al. reported that SIRT1 can prevent heart aging by mediating deacetylation of PGC-1α and activation of mitochondrial biogenesis (Wang et al., 2020a). To summarize, morphological and quantitative changes in mitochondria have been observed in the heart, liver, and skeletal muscle in pathological conditions, and the regulation of mitochondrial biogenesis is a potential target to treat mitochondrial dysfunction.

Mitochondrial dynamics, represented by the balance of mitochondrial fission and fusion, are under the regulation of dynamin-related guanosine triphosphatase (GTPase) proteins (Hoppins et al., 2007). Dynamin-related protein 1 (DRP1) is the main protein of fission, which interacts with the receptors located in OMM, such as mitochondria fission factor (MFF), fission protein-1 (Fis1), and mitochondrial dynamics proteins 49 and 51 (MiD49/51) (Long et al., 2013; Richter et al., 2014; Kornfeld et al., 2018). Besides, MiD49/51 and MFF have also been reported to modulate GTPase-derived DRP1 constrictive activity (MiD49/51 inhibits GTPase, whereas, MFF promotes it) (Liu and Chan, 2015; Osellame et al., 2016). During the fission phase, dysfunctional fragments of mitochondria can be cut into small, spherical pieces and then degraded by mitophagy. Excessive mitochondrial fission is responsible for increasing the infarct area in I/R heart injury, and it induces both excessive mitophagy and cell death (Bi et al., 2019). Drugs such as metformin can block DRP1 and maintain the integrity of mitochondria according to the studies using diabetic animal models in vivo (Li et al., 2016; Wang et al., 2019). Recently, metformin is also demonstrated to protect against AS by reducing cardiac mitochondrial fission (it is most well recognized by its mechanism in the regulation of mitochondrial complex I, mainly through the AMPK pathway, discussed in 3.3) (Wang et al., 2017). Mitochondrial fusion is first mediated by mitofusin 1 (Mfn1) and Mfn2 in the OMM, and then by optic atrophy 1 (Opa1) in the IMM (Koshiba et al., 2004). Deletion of Mfn1 and Mfn2 in the embryo causes death in utero, indicating that Mfn1 and Mfn2 are strongly contributed to heart development (Chen et al., 2003; Chen et al., 2011). There is evidence from murine hearts that the ablation of Mfn1 and Mfn2 is related to the fragmentation and malfunction of mitochondria and induction of lethal cardiomyopathy (Papanicolaou et al., 2012). A decreasing level of Opa1 is found in post-MI hearts of both rats and humans and correlates with the downregulation of mtDNA and antioxidant genes (Chen et al., 2009). Mitochondrial fusion is a beneficial process in which mitochondria with tiny defects of proteins and DNA are complemented by functional parts of other mitochondria, and are merged into elongated, tubular, interconnected mitochondrial networks (Campello and Scorrano, 2010).

Mitophagy is a form of autophagy that selectively targets the degradation of dysfunctional mitochondria and serves as a control system of mitochondria, and autophagosomes are the main effectors during mitophagy (Lemasters, 2005). Together with mitochondrial biogenesis and dynamics, mitophagy sustains the homeostasis of mitochondria. At the cellular level, mitophagy shows effects on cell death of CMs (Shires and Gustafsson, 2015; Torrealba et al., 2017), and regulates the differentiation of cardiac fibroblasts (CFs), the main responding cell population in post-MI hearts, thereby controlling cardiac remodeling (Urban et al., 2015). Normal levels of mitophagy protect against I/R damage. Several studies have investigated that repeated and short iPre promotes the adaptation of cardiac mitochondria to I/R injury via mitophagy (Zhang et al., 2016; Zhang et al., 2017; Xiao et al., 2020). In addition, mitophagy contributes to the remission of other CVDs, such as hypertension, cardiac remodeling after MI and diabetic cardiomyopathy (more details are provided in the review by Morales et al.) (Morales et al., 2020). Several important pathways are involved in different phases of mitophagy. Mitophagy usually starts with mitochondrial imperfection and a reduction in mitochondrial membrane potential (Δψm). Regularly, PTEN-induced kinase 1 (PINK1) recognizes damaged mitochondria and accumulates in the OMM, thereby leading to the phosphorylation of Parkin E3 ligase. Parkin helps to ubiquitinate proteins of the OMM (Kondapalli et al., 2012). Mitochondrial cargo receptors (MCRs) guide these proteins to bind with microtubule-associated protein 1A/1B-light chain 3 (LC3), which can be recognized by autophagosomes and finally digested (Lazarou et al., 2015). Interestingly, DRP1 can also help selectively recognize damaged mitochondria from healthy population by mitochondrial fission. PINK sometimes cannot recognize the damaged part of mitochondrion easily due to the interconnected rapid balance of electrochemical gradients. However, transient reduction of Δψm can be a negative selective pressure for DRP1-induced fission. Healthy part of the mitochondrion will enter the fission-fusion cycle. Whereas, dysfunctional sectors which cannot recover back to normal Δψm can be negatively selected by mitophagy. Additionally, DRP1, together with Δψm, can also induce a positive feedback loop (reduction of Δψm recruits DRP1, recruited DRP1 ulteriorly reduces Δψm), which is helpful to selective mitophagy. (Cho et al., 2019). Other MCRs can directly stimulate mitophagy in ubiquitination-independent pathways. Cardiolipin is mostly synthesized on the IMM (Ren et al., 2014) and has effects on the respiratory chain and mitochondrial quality control system (mitochondrial dynamics and mitophagy) (Dudek, 2017; Paradies et al., 2019). Damage signals of mitochondria cause the translocation of cardiolipin to the OMM, which combines with LC3 and contributes to the recognition of dysfunctional (part of) mitochondria in the heart (Chu et al., 2013; Dudek et al., 2019). Furthermore, B-cell lymphoma-2 (BCL2)-interacting protein 3 (BNIP3) and FUN14 domain containing 1 (FUNDC1) have been reported to activate mitophagy via a PINK1-independent pathway and help to protect against oxidative injury and I/R in the heart (Liu et al., 2012; Zhang et al., 2016).

Based on the mechanisms of mitochondrial injury/repair mentioned above, we summarize mito-targeted modulators against mitochondrial dysfunction in CVDs.

The heart is very dependent on the normal functioning of mitochondria due to its high energy consumption and ATP-sensitive characteristics. Scientists are becoming increasingly aware of the importance of mitochondrial targets for the treatment of CVDs. The topic is not only about the investigations of new targets and their mechanisms, but also about the delivery systems of these drugs. The targets against mitochondrial dysfunction mainly focus on antioxidants, repair of the respiratory chain, and mitochondrial homeostasis (including mitochondrial biogenesis, mitochondrial dynamics, and mitophagy). Furthermore, scientists have gradually accepted that ncRNAs play important regulatory roles in cardiac mitochondria, and future targeted therapies based on ncRNA may become a hot topic (all the targets mentioned in this review and their evidence are summarized in Table 1). However, there remain some issues related to mitochondria-targeted therapies. Firstly, some of the compounds have less effect on dysfunctional mitochondria than expected. For example, TPP⁺-linked modulators (such as MitoQ, MitoTEMPO, and SkQ1) can combine with the OMM due to their lipophilic characteristics, which rely on the accumulation of cationic molecules via Δψm (Jiang et al., 2009). However, Δψm is decreased in dysfunctional mitochondria (Zorova et al., 2018), which may reduce the therapeutic efficiency of these targets. Secondly, other targets are presenting accompanying effects during global treatment (tissue-unspecific treatment), the MPTP inhibitor, CsA, representing a good example. Namely, CsA, has shown cardioprotective function in different species/models (Onishi et al., 2012; Badavi et al., 2014; González Arbeláez et al., 2016; Lozano et al., 2020), however, CsA has a significant immunosuppressive effect at the same time (Frank et al., 1999). To solve the issues of low targeting efficiency and accompanying effects, pharmacologically specific targets such as NIM811, which is also an inhibitor of MPTP but has no immunosuppressive effect compared with CsA (Argaud et al., 2005), warrant further investigation. Another way is to establish a more precise way to transport compounds into the heart. Ikeda et al. synthesized nanoparticle-mediated CsA using poly-lactic/glycolic acid (PLGA) nanoparticles, which target injured organs and have shown promising effects on mouse heart models (Ikeda et al., 2016). Therapies based on nanoparticle-mounted targets (even intervention with small-interference RNA of ncRNA) may become a new direction for the treatment of mitochondrial dysfunction and CVDs. Furthermore, the impact of sterile inflammation and immunity induced by damaged mtDNA in CVDs cannot be ignored. The main solution now is to protect mtDNA against ROS damage. Besides, promising PRRs like NLRPs have been well-studied in animal models as a potential target involving inflammation response in CVDs (Molagoda et al., 2021). Clinical drugs, such as metformin and proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor, also have the potential to suppress NLRP3-regulated inflammation in the heart (Wang et al., 2020b; Fei et al., 2020). However, targets towards mtDNA and inflammation response involve a vast amount of inflammatory/immune cytokines and pathways. The major problems at present are finding more efficient co-/interventional targets and specific drug delivery systems.